2060
YOL.
NOTES
was established by analysis of the diamine (I) for primary amino nitrogen which was absent. The purity of the diamine (I) was ultimately established by its polymerization with polytetramethyleneether glycol bischloroformate to form a high molecular weight polyurethane.' This sequence of reactions should provide a general route for the preparation of unsymmetrically N,h-'-disubstituted ethylenediamines of high purity. 0 0
/I
I!
CH?OC--COCHa
I1
+ HNCH2CHZNH2 4 i I
CH, CH2
i
OH I11
0
0
C-
c
!I
I/
HOCHzCHsN/
N ' -H
\
CHz-CH, IV
&He
I CH,
CH,
CH2
OH
CHz
I
/
I
I
i
I II
I
CH
/I
I
CH:. 0
1
0 II
EXPERIMENTAL
N-(%HydroxyethyL)piperazine-$,3-dione (IV). A mixture of 343 g. (3.30 moles) of aminoethylethanolamine (111) and 5 ml. of conrd. hydrochloric acid was added to 389 g. (3.30 moles) of dimethyl oxalate (11) over a period of 15 min. with good agitation. The temperature was raised gradually to 218' in about 1 hr., during x-hich time 206 g. (6.44 moles, 97.6%) of methanol distilled from the reaction mixture. The reactants gradually formed a viscous polymer which broke down above 180" to form the piperazinedione (IV) and a red noncrystalline material which was not further investigated. The mixture was cooled to room temperature, taken up in 400 ml. of ethanol, cooled, and filtered. The crude product (199 g., 38%) was recrystallized from alcohol until pure, m.p. 163-164". Anal. Calcd. for C6H1J203: C, 45.56; H, 6.37; S, 17.72. Found: C, 45.2, 45.5; H, 5.9, 6.1; N, 17.4, 17.6. N-(RH ydroxyethzJZ)-.V'-(d-penten y l ) p i p e r a z i n e d ,3-dione ( V ) . h 3-l., four-necked flask equipped with a stirrer, thermometer, and reflux condenser fitted with a calcium sulfate drving tube xas flamed out and cooled while being flushed n-ith dry nitrogen. Distilled t-butyl alcohol (1600 ml.) was added followed by 61.4 g. (1.57 moles) of potassium. The mixture was refluxcd and agitated until the metal had com-
25
pletely reacted. N-( %Hydroxyethyl)piperazine-2,3-dione (248g., 1.57 moles) was added, and the agitated suspension was refluxed overnight. The temperature was lowered to 70°, 234 g. (1.57 moles) of 1-bromo-4pentene3 was added, and the mixture was again refluxed overnight. After cooling, the solid potassium bromide was filtered (160 g., 86Ojo), and the t-butyl alcohol wae distilled. The last traces of solvent were removed under reduced pressure. The viscous residue was extracted with benzene (one 500-ml. and three 250-ml. portions) and then with tetrahydrofuran (three 500ml., eight 250-ml., and six 100-ml. portions). The tetrahydrofuran was distilled, and the residue placed in a 0' coldbox overnight to crystallize. The solid was recrystallized from tetrahydrofuran (wt. 127 g.). The filtrate was diluted with 2 1. of tetrahydrofuran and the supernatant liquid was decanted from the precipitated oil. The tetrahydrofuran solution was again concentrated, seeded, and cooled to yield another 56 g. of solid. Further concentration of the filtrate yielded an additional 8.5 g. of product, bringing the total yield to 191.5 g. (54y0).The residue from the benzene extract, combined with the end tetrahydrofuran filtrate from the recrystallizations, was chromatographed on 200-mesh activated alumina with tetrahydrofuran and ethanol. This resulted in the recovery of an additional 56.5 g. (16%) of material. Recrystallization of the combined solids from tetrahydrofuran afforded pure AT-(2-hydroxyethyl)-N'-(4pentenyl)piperazine-2,3dione, m.p. 75-76.5'. Anal. Calcd. for C11H18N203: C, 58.39; H, 8.02; N, 12.38. Found: C, 58.5, 58.5; H, 7.8, 7.9; N, 12.5, 12.7.
N-(,%'-Hydroxyet?ayl)-N'-(4-pentenyl)ethylenediamine (I).TO a solution of 50 g. (0.770 mole) of 85% potassium hydroxide in 500 ml. of ethanol was added 0.1 g. of 2,6-di-t-butyl-pcresol, a solution of 0.1 g. of sodium sulfite in 25 ml. of distilled water, and 83 g. (0.376 mole) of A7-(2-hydroxyethyl)N'-(4pentenyl)piperazine-2,3-dione ( V). The clear solution Tas refluxed overnight under an atmosphere of nitrogen. A precipitate began forming after about 15 min. The mixture was cooled, the solid potassium oxalate monohydrate was filtered (65.5 g., 96.8%), and the solvent was distilled from the filtrate. Vacuum distillation of the residue yielded hydroxyethylpentenylethylenediamine(57.6 g., 9 1.5% yield, b:p. 97.5' (0.i5 mm:), n e 1.4772). Anal. Calcd. for CgH&20: C, 62.75; H, 11.70; X, 16.27; urimarv amino N. absent. Found: C. 62.5, 62.8; H, 11.4, i1.5; N, 16.1, 16.1; primary amino N; absent. A drop of amine added to aqueous oxalic acid yielded the bisoxalate, m.p. 236-23'7'. Anal. Calcd. for C,aHz,N,Og: C, 44.31; H, 6.87; N, 7.95. Found: C, 44.0, 44.2; H, 6.8, 6.9; N, 7.8, 7.8. The amine forms a solid hemihydrate, m.p. 41.5-42", on admixture with 0.5 mole equivalent of water.
ELASTOMER CHEMICALS DEPARTMENT E. I. DU PONT DE XEMOURS AND Co., INC. WILMINGTON, DEL. (3) Prepared according to the method of P. Gaubert, R. P. Linstead, and H. N. Rydon, J . Chem. SOC.,1971 (1937) and E. M. Van Heyningen, J . Chem. SOC.,76, 2241 (1954), b.p. 124' (760 mm.), n y 1.4615 (reported b.p. 124.5-128", n y 1.4642).
Decarboxylation of N-Methylaminoaromatic ortho-Carboxylic Acids WAYLAND E. NOLAND AND GEORGE J . MEISTERS~ Received A p r i l 7', 1960
Dry distillation with soda-lime has been shown previously to produce AT-demethylation (and con-
KOVEMBER
1960
comitant decarboxylation to carbazole) of the salt of 9-methylcarbazole-l,2,3,4-tetracarboxylic acid (I). No N-demethylation occurred under similar conditions, however, with 9-methylcarbazole2 or with the salt of a triacid believed to be l-methylindole4,5,6-tricarboxylic acid (11). 3 The possibility that
I.::n:x""" COOH
COOH
h-
COOH HOOC H O y Q
CH, COOH
CH3
I
I1
IT-demethylation of I is occurring by neighboring group participation of the ortho carboxylate anion led us to test the generality of N-demethylation accompanying decarboxylation by use of selected N-methylaminoaromatic ortho carboxylic acids. Ar-Methylanthranilic acid (111),4 upon decarboxylation with soda-lime,2 gave a liquid amine, which was isolated in 34% yield as the acetyl derivative, shown to be identical by mixed melting point and infrared comparison in Nujol with an authentic sample of N-methylacetanilide. !+Methylcarbazole-1-carboxylic acid (IV)5 and 9-methylcarbazole- 1,8-dicarboxylic acid (V) were similarly decarboxylated to 9-methylcarbazole in yields of 587, and 507,, respectively.
I11
&H,COOH HOOC
6~~COOH
IV
V
These results show that N-demethylation accompanying decarboxylation of N-methylaminoaroniatic ortho carboxylic acid salts is not a general phenomenon; that the presence of one or two ortho carboxylate groups is, in itself, insufficient cause for N-demethylation. It is concluded that, in the case in which it occurs (I),N-demethylation is favored by unusual stabilization of the resulting anion, attributable to the combined resonance and inductive effects of the four carboxylate substituent groups. In this connection, it is perhaps of interest to note that our efforts to Ar-methylate the sodium salt of the ester ( V I ) 2of the corresponding uiisubstituted COOCH3 COOCHs
@ \
_
_
~
.
N
COOCH, COOCH, VI
(1) GradLate School research assistant, summer 1959.
It is a pleasure to acknowledge the support of this work through a grant from the General Research Fund of the Graduate School of the University of Minnesota. (2) W. E. Noland, TV. C. Kuryla, and R. F. Lange, J . Am. Chem. SOC.,81, 6010 (1959). (3) 0. Diels, K. Alder, and H. Winckler, Ann., 490, 267 ( 1931).
2061
NOTES
compound have been unsuccessful. The results cited in this paper provide no evidence either for or against neighboring group participation by the ortho carboxylate group in N-demethylation in the case in which it occurs.2 Because of its importance to the conclusions drawn in this work, it appeared desirable to establish rigorously the position of the carboxyl group in 9-methylcarbazole-1-carboxylic acid (IV), prepared by action of n-butyllithium on 9-methylcarbazole, followed by reaction with carbon d i ~ x i d eThe . ~ structure had been logically assigned previously by analogy with numerous examples of ortho lithiation of aromatic amines5 and with the fact that 9-ethylcarbazole-1-carboxylic acid, prepared similarly, had been shown to be identical with a sample prepared by ethylation of carbazole-1-carboxylic acid.6 Carbazole-1-carboxylic acid, most readily prepared from carbazole by reaction of its potassium salt' with carbon dioxide at 270°,s has been prepared unambiguously by two different method^,^, lo and the product in the first case has been shown to be identical with a sample prepared from carbazole. After unsuccessful methylation attempts with dimethyl sulfate or methyl iodide in the presence of alkali, l 1 we obtained 9-methylcarbazole-1-carboxylic acid (IV) in 11% yield from carbazole-lcarboxylic acid8 by the action of sodamide and methyl iodide in liquid ammonia.12The sample was shown to be identical by mixed melting point (188.5-190') and infrared comparison in Nujol with a sample6 prepared from 9-methylcarbasole, thus constituting a proof of structure for 9-methylcarbazole-1-carboxylic acid (IV). EXPERIMENTAL
Melting points were determined on a calibrated Kofler micro hot stage. Decarboxylation of N-methylanthranilic acid (111). N Methylanthranilic acid4 (3.00 g., 0.0198 mole) was mixed thoroughly with powdered soda-lime (9 g.) and the mixture pyrolyzed under a stream of nitrogen with the aid of a Meker burner, in the manner previously described.* The liquid distillate was refluxed m
